Selection bias in gene extraction on the basis of microarray gene-expression data

In the context of cancer diagnosis and treatment, we consider the problem of constructing an accurate prediction rule on the basis of a relatively small number of tumor tissue samples of known type containing the expression data on very many (possibly thousands) genes. Recently, results have been presented in the literature suggesting that it is possible to construct a prediction rule from only a few genes such that it has a negligible prediction error rate. However, in these results the test error or the leave-one-out cross-validated error is calculated without allowance for the selection bias. There is no allowance because the rule is either tested on tissue samples that were used in the first instance to select the genes being used in the rule or because the cross-validation of the rule is not external to the selection process; that is, gene selection is not performed in training the rule at each stage of the cross-validation process. We describe how in practice the selection bias can be assessed and corrected for by either performing a cross-validation or applying the bootstrap external to the selection process. We recommend using 10-fold rather than leave-one-out cross-validation, and concerning the bootstrap, we suggest using the so-called. 632+ bootstrap error estimate designed to handle overfitted prediction rules. Using two published data sets, we demonstrate that when correction is made for the selection bias, the cross-validated error is no longer zero for a subset of only a few genes.

Self-Similarity of Air-Water Flows in Skimming Flows on Stepped Spillways

Skimming flows on stepped spillways are characterised by a significant rate of turbulent dissipation on the chute. Herein an advanced signal processing of traditional conductivity probe signals is developed to provide further details on the turbulent time and length scales. The technique is applied to a 22° stepped chute operating with flow Reynolds numbers between 3.8 and 7.1 E+5. The new correlation analyses yielded a characterisation of large eddies advecting the bubbles. The turbulent length scales were related to the characteristic depth Y90. Some self-similar relationships were observed systematically at both macroscopic and microscopic levels. These included the distributions of void fraction, bubble count rate, interfacial velocity and turbulence level, and turbulence time and length scales. The self-similarity results were significant because they provided a picture general enough to be used to characterise the air-water flow field in prototype spillways.

ESEM observations of SCC initiation for 4340 high strength steel in distilled water

The stress corrosion cracking (SCC) initiation process for 4340 high strength steel in distilled water at room temperature was studied using a new kind of instrument: an environmental scanning electron microscope (ESEM). It was found that the applied stress accelerated oxide film formation which has an important influence on the subsequent SCC initiation. SCC was observed to initiate in the following circumstances: (1) cracking of a thick oxide film leading to SCC initiation along metal grain boundaries, (2) the initiation of pits initiating SCC in the metal and (3) SCC initiating from the edge of the specimen. All these three SCC initiation circumstances are consistent with the following model which couples SCC initiation with cracking of a surface protective oxide. There is a dynamic interaction between oxide formation, the applied stress, oxide cracking, pitting and the initiation of SCC. An aspect of the dynamic interaction is cracks forming in a protective surface oxide because of the applied stress, exposing to the water bare metal at the oxide crack tip, and oxidation of the bare metal causing crack healing. Oxide crack healing would be competing with the initiation of intergranular SCC if an oxide crack meets the metal surface at a grain boundary. If the intergranular SCC penetration is sufficiently fast along the metal grain boundary, then the crack yaws open preventing healing of the oxide crack. If intergranular SCC penetration is not sufficiently fast, then the oxidation process could produce sufficient oxide to fill both the stress corrosion crack and the oxide crack; in this case there would be initiation of SCC but only limited propagation of SCC. Stress-induced cracks in very thin oxide can induce pits which initiate SCC, and under some conditions such stress induced cracks in a thin oxide can directly initiate SCC.

Chemical treatment of Escherichia coli. II. Direct extraction of recombinant protein from cytoplasmic inclusion bodies in intact cells

A method is presented for the direct extraction of the recombinant protein Long-R-3-IGF-I from inclusion bodies located in the cytoplasm of intact Escherichia coli cells. Chemical treatment with 6M urea, 3 mM EDTA, and 20 mM dithiothreitol (DTT) at pH 9.0 proved an effective combination for extracting recombinant protein from intact cells. Comparable levels of Long-R-3-IGF-I were recovered by direct extraction as achieved by in vitro dissolution following mechanical disruption. However, the purity of directly extracted recombinant protein was lower due to contamination by bacterial cell components. The kinetics of direct extraction are described using a first-order equation with the time constant of 3 min. Urea appears important for permeabilization of the cell and dissolution of the inclusion body. Conversely, EDTA is involved in permeabilization of the cell wall and DTT enhances protein release. pH proved to be important with lower levels of protein release achieved at low pH values (

The effect of water fluoridation on the bone mineral density of young women [Saskatoon study]

Introduction: Osteogenic effects of therapeutic fluoride have been reported; however, the impact of exposure to low level water fluoridation on bone density is not clear. We investigated the effect of long-term exposure to fluoridated water from growth to young adulthood on bone mineral density (BMD). Methods: BMD was measured in 24 healthy women from Regina (fluoride 0.1 mg/L) and 33 from Saskatoon (fluoride 1.0 mg/L), with no differences between groups for height, weight, lifestyle or dietary factors. Results: Saskatoon women had significantly higher mean BMD at total anterior-posterior lumbar spine (APS) and estimated volumetric L3 (VLS), with no difference at total body (TB) or proximal femur (PF). Conclusion: Exposure to water fluoridation during the growing years may have a power impact on axial spine bone density in young women.

Free volume and water uptake in a copolymer hydrogel series

A copolymer of X-hydroxyethyl methacrylate (HEMA) with 2-ethoxy ethyl methacrylate (EEMA) was synthesized and the molecular mobility, free volume, and density properties examined as a function of composition. These properties were correlated with the equilibrium water uptake in order to determine which of the properties were most influential in causing high water sorption, as these materials are suitable candidates for hydrogel systems. It was found that the polar HEMA repeat unit results in a rigid, glassy sample at room temperature due to the high degree of hydrogen bonding between chains whereas high EEMA content leads to rubbery samples with subambient glass transition temperatures. The free volume properties on the molecular scale measured by positron annihilation Lifetime spectroscopy (PALS) showed that higher HEMA content led to smaller, fewer holes and a lower free volume fraction than EEMA. Therefore the high water uptake of HEEMA-containing copolymers is largely related to the high polarity of the HEMA unit compared to EEMA, despite the low content of free volume into which the water can initially diffuse. Trends in density with copolymer composition, as measured on a macroscopic level, differs to that seen by PALS and indicates that the two techniques are measuring different scales of packing. (C) 1998 John Wiley & Sons, Inc.

Bioimpedance spectrometry in the determination of body water compartments: Accuracy and clinical significance

Bioelectrical impedance analysis (BIA) offers the potential for a simple, portable and relatively inexpensive technique for the in vivo measurement of total body water (TBW). The potential of BIA as a technique of body composition analysis is even greater when one considers that body water can be used as a surrogate measure of lean body mass. However, BIA has not found universal acceptance even with the introduction of multi-frequency BIA (MFBIA) which, potentially, may improve the predictive accuracy of the measurement. There are a number of reasons for this lack of acceptance, although perhaps the major reason is that no single algorithm has been developed which can be applied to all subject groups. This may be due, in part, to the commonly used wrist-to-ankle protocol which is not indicated by the basic theory of bioimpedance, where the body is considered as five interconnecting cylinders. Several workers have suggested the use of segmental BIA measurements to provide a protocol more in keeping with basic theory. However, there are other difficulties associated with the application of BIA, such as effects of hydration and ion status, posture and fluid distribution. A further putative advantage of MFBIA is the independent assessment not only of TBW but also of the extracellular fluid volume (ECW), hence heralding the possibility of,being able to assess the fluid distribution between these compartments. Results of studies in this area have been, to date, mixed. Whereas strong relationships of impedance values at low frequencies with ECW, and at high frequencies with TBW, have been reported, changes in impedance are not always well correlated with changes in the size of the fluid compartments (assessed by alternative and more direct means) in pathological conditions. Furthermore, the theoretical advantages of Cole-Cole modelling over selected frequency prediction have not always been apparent. This review will consider the principles, methodology and applications of BIA. The principles and methodology will,be considered in relation to the basic theory of BIA and difficulties experienced in its application. The relative merits of single and multiple frequency BIA will be addressed, with particular attention to the latter's role in the assessment of compartmental fluid volumes. (C) 1998 Elsevier Science Ltd. All rights reserved.

Improved one-step solid-phase extraction method for morphine, morphine-3-glucuronide, and morphine-6-glucuronide from plasma and quantitation using high-performance liquid chromatography with electrochemical detection

This communication describes an improved one-step solid-phase extraction method for the recovery of morphine (M), morphine-3-glucuronide (M3G), and morphine-6-glucuronide (M6G) from human plasma with reduced coextraction of endogenous plasma constituents, compared to that of the authors' previously reported method. The magnitude of the peak caused by endogenous plasma components in the chromatogram that eluted immediately before the retention time of M3G has been reduced (similar to 80%) significantly (p < 0.01) while achieving high extraction efficiencies for the compounds of interest, viz morphine, M6G, and M3G (93.8 +/- 2.5, 91.7 +/- 1.7, and 93.1 +/- 2.2%, respectively). Furthermore, when the improved solid-phase extraction method was used, the extraction cartridge-derived late-eluting peak (retention time 90 to 100 minutes) reported in our previous method, was no longer present in the plasma extracts. Therefore the combined effect of reducing the recovery of the endogenous components of plasma that chromatographed just before the retention time of M3G and the removal of the late-eluting, extraction cartridge-derived peak has resulted in a decrease in the chromatographic run-time to 20 minutes, thereby increasing the sample throughput by up to 100%.